This submission by the Canadian Bankers Association is in response to the Government of Canada's June 23, 2003 Consultation Paper, specifically regarding the issue of expanding the functionality of automated banking machines. This submission canvasses a wide range of issues the banking industry believes policy makers should take into account as they consider the implications of broadening the shared functionality of ABMs. These issues are examined in light of the possible impact on competition in the marketplace, customer service, risk management, and institutional operations (including a discussion of important differences between individual and commercial accounts). This submission also surveys relevant developments in other leading countries and canvasses the operational and public policy implications of expanding functionality. Further, the submission stresses the need to define what is meant by "full functionality", particularly since the Consultation Paper describes full functionality as "deposits, withdrawals, bill payments, etc.", leaving its parameters seemingly open-ended and undefined and including functions (e.g., bill payments) that are qualitatively different than other financial services transactions.

In the detailed discussion of these matters in this submission, three key considerations stand out:

No Mandatory Functionality:

The threshold public policy issue raised by the Consultation Paper is whether ABM functionality should continue to be a voluntary matter for individual institutions to consider in response to customer demand, competitive pressures, marketplace conditions, and their individual business plans or whether the Government should mandate expanded shared functionality. While individual bank financial groups may have different views on the merits and feasibility of expanding functionality, they share the common position that increased regulation and government imposition of mandatory shared functionality would not be in the public interest. Shared ABM functionality is already available to Canadians today through two different networks – Interac (for shared cash dispensing) and The Exchange Network (for shared cash dispensing, deposits, account transfers, and balance inquiries). It is important to note in this regard that other leading jurisdictions – such as Australia and the UK – do not mandate shared functionality, and that the clear trend in the US is away from mandated shared functionality in favour of allowing the marketplace to respond to competitive pressures.

Effect on Competition, Customer Service, and Innovation:

It is not clear that mandating expanded functionality would achieve the government's stated objective of enhancing competition. While it is the view of some institutions that expanding ABM functionality could have a positive impact on competition by expanding the number of delivery channels available to smaller players, other institutions take a very different view. They believe that mandatory expanded functionality could have a range of negative effects, including a decline in the number of ABMs, reduced customer service levels and investment in ABM technology, and reduced innovation because institutions would face disincentives to enhancing their networks. These banks view their proprietary ABM networks as one important way to differentiate themselves from each other and from their non-bank competitors in the financial services marketplace, and as such a source of competitive advantage that has merited significant investment in innovation and expansion.

If mandated shared functionality is intended to effectively increase competition, there should be clear evidence that it would be widely adopted by consumers and small business. The industry in unaware of any such evidence and survey results included in this submission indicate the opposite may be the case. In addition, an issue that would need to be seriously considered is whether holds on ABM deposits, a customer concern today, would increase under a shared deposit arrangement.

Operating Considerations and Costs of Shared Functionality:

Another key consideration is the extent to which expanding functionality would reduce operational efficiency and increase costs. The consequences of grafting manual processing requirements onto a largely electronic system would be significant. In The Exchange Network, shared deposit taking involves a number of manual steps including: reconciling the contents of the deposit envelopes to the amounts of the deposit transactions; notifying the Issuer of discrepancies; endorsing cheques in the deposit envelopes; and photocopying the fronts and banks of cheques over $5000 and faxing the copies to the Issuer.

At current volume levels (e.g., one large Exchange Network member reports an average of 60 shared deposits per month), the Exchange processes for deposit reconciliation appear to require little, if any, additional staffing resources to verify and reconcile deposits. However, this would not be the case if larger banks with their much larger transaction volumes (nearly one million deposits per month across the system) were required to provide shared deposit functionality using processes similar to that the Exchange Network. If some or all of the above processes had to be put in place, the staffing requirements and impacts on operational efficiency would be significant. These impacts would be compounded considerably if commercial accounts – which typically have much larger volumes and larger dollar amounts than retail accounts – were included in mandatory shared deposit taking. Given the possible extent of the costs, full recovery of the costs applied at the transaction level would likely be unattractive to customers. Further research and consideration would be needed to fully assess the impacts and changes that would be needed in this area, and what the resulting costs to consumers would be.

These and other issues addressed in the submission make clear that a shift in current policy would have significant implications for the future direction of the Canadian banking industry, non-bank competitors, competing ABM networks, and for consumers of financial products and services. Policy makers should give careful consideration to each of the issues and questions the banking industry identifies in this canvass of the possible impact of expanded shared ABM functionality.

1. Introduction

The Canadian Bankers Association welcomes the opportunity to comment on the Government of Canada's June 23, 2003 response to the reports of the House of Commons Standing Committee on Finance and the Standing Senate Committee on Banking, Trade and Commerce regarding large bank mergers and the public interest. Specifically, this submission responds to the public policy issues raised in the government's paper regarding expanding the functionality of automated banking machines ("ABMs").

Our submission canvasses a wide range of issues we believe policy makers should take into account as they consider the implications of broadening the shared functionality of ABMs. These issues are examined in light of the possible impact of expanded functionality on competition in the marketplace, customer service, risk management, and institutional operations (including a discussion of important differences between individual and commercial accounts). In addition, this submission surveys relevant developments in other leading countries and canvasses the operational and public policy implications of expanding functionality. In developing this paper, the CBA conducted a detailed study of the Exchange Network, a national network consisting of over 1,300 ABMs sharing four services over the proprietary networks of its 68 participating financial institutions, and some of the observations in this paper are drawn from our analysis of that network.

The following key points preface our submission.

a) No Increased Regulation

Canadians have access to a wide range of competitive and innovative ABM networks in Canada – both shared and proprietary – established by competing deposit-taking institutions, and indeed are the highest per capita users of ABM services in the world with 47.8 ABM transactions per Canadian in 2001. The growth and development of these networks has been driven primarily by institutions' competitive responses to market forces and consumer demand. As the issue of expanding functionality is reviewed, it is critical that policy-makers fully take into account the importance of these drivers of continuing innovation. Any mandatory imposition of expanded functionality would not be a positive or welcome development.

b) ABM Networks Differ From Interac

Bank ABM delivery channels should not be equated with Interac's shared cash dispensing service. While banks have voluntarily agreed to offer shared cash dispensing to their customers through Interac, a significant majority of transactions performed by customers on a bank ABM are not related to Interac's services. For example, in the case of cash withdrawals, only those from an ABM not belonging to a cardholder's financial institution are Interac transactions. Having regard to the largest banks, approximately 70 percent of cash withdrawals are 'on us' withdrawals (i.e., non-Interac). Moreover, other functions offered to bank customers at proprietary ABMs such as deposit taking, account updating, cheque ordering, account transfers, and bill payments are non-Interac transactions. In other words, while Interac is a shared system that institutions use for certain services (and while some deposit-taking institutions, including some Interac members, have joined together to offer a range of additional shared services through the Exchange Network), proprietary ABMs are considered by many deposit-taking institutions to primarily be competitive delivery channels for their own customers.

c) Need to Define "Full Functionality"

It is essential to understand what is meant by "full functionality" before considering the appropriate policy responses to the question. Testimony on this issue before the two parliamentary committees referred only to shared deposits and withdrawals in the context of expanding functionality. The government's response paper, however, suggests a broader meaning, describing full functionality as "deposits, withdrawals, bill payments, etc.", leaving its parameters seemingly open-ended and undefined. It bears noting that the shared functionality offered by the Exchange Network (a voluntary shared ABM network currently operating in Canada that serves as a competitive response for some deposit-taking institutions to larger players in the market) consists of deposits, withdrawals, transfers and balance inquiries. A more precise understanding of the intended breadth of expanded ABM functionality is necessary to assess all of the potential implications of a policy change in this area. For the purposes of this submission, when considering the public policy and operational implications of broadening or expanding the shared functionality of ABMs, we refer, not to "full functionality", but to "expanded" or "shared" functionality.

Further, a more precise understanding is also needed of which providers would be affected if policy changes were to be made by the federal government in this area. For instance, would provincially-regulated credit unions and caisses populaires be included? Further, would "white-label" ABM providers be included, and if so, would these providers be required to convert their machines to become deposit-takers? Similarly, would regulated financial institutions whose mix of ABMs include machines that only dispense cash, be required to reconfigure these machines to provide services not currently available even to the provider's own customers?

d) Enhancement of Competition

It is not clear that mandating expanded functionality would achieve the government's stated objective of enhancing competition. While it has been suggested by some financial institutions, as discussed in more detail below, that expanding ABM functionality would have a positive impact on competition, other institutions take a very different view. They believe that mandatory expanded functionality would have a negative effect on the market, particularly regarding innovation and expansion of the ABM network. For instance, they note that there is no evidence in the ABM marketplace of consumer demand for expanded functionality. To the contrary, a 2003 survey, conducted by The Strategic Counsel for Interac, shows a small and declining support over the past three years for expanding functionality beyond shared cash dispensing. Indeed, only 12% of respondents were interested in making deposits at an ABM belonging to another financial institution, while only 6% were interested in shared balance inquiries and 3% in shared account transfers. Similarly, as discussed in greater detail below, it does not appear that small business would desire expanded functionality. Thus, in the absence of any apparent customer demand for expanded functionality, and with clear evidence of strong differences of view regarding impacts on competition, careful study would be required before implementing the fundamental changes to present ABM systems that expanded functionality would require.

2. Competition in the Marketplace

The questions posed by the government in its response to the parliamentary committee reports focus on the possible competitive impacts of broadening the shared functionality of ABMs. Indeed, some deposit-taking institutions believe that expanding the functionality of ABMs would add a new element of competition in the financial services sector by expanding the number of delivery channels available to these institutions, and thereby strengthening their offer to customers. Others see expanded functionality leading to a decline in the number of ABMs, reduced service levels and investment in ABM technology because institutions face disincentives to enhancing their networks. Accordingly, we canvass first the possible effects of changes in this area on competition in the financial services marketplace.

When considering the possible competitive implications of expanding shared ABM functionality, one important observation needs to be made at the outset. Shared ABM functionality is available to Canadians today through two different networks. The members of Interac have agreed voluntarily to provide access to non-customers at their respective ABMs networks for certain services. Interac makes its services available to members, but it is important to note that its members retain the choice to decide which of these services (e.g., cash dispensing or point-of-sale services or both) to offer, and also whether to be an Acquirer, an Issuer, or both. The second shared network is (as noted above) the Exchange Network. Provided that they meet the Exchange's membership criteria, any bank or other deposit-taking institution can become a member of this network. The existence of Interac and the Exchange Network demonstrate the capacity of the Canadian financial services marketplace to respond in its own way to competitive needs driven by consumer demand.

The following two issues arose in our canvass of the competitive impact of broadening shared ABM functionality that should be taken into account by policy makers.

a) Differing Strategic Approaches

ABMs are one of many delivery channels that financial institutions have developed for use by their customers. Individual bank investments in their respective ABM delivery channels are a direct result of strategic business planning. While some deposit taking institutions, mainly large banks, have chosen to invest extensively in proprietary ABMs, some smaller institutions have decided to have relatively fewer machines and rely more heavily on sharing services with other parties with whom they contract (e.g., through the Exchange Network), while still others have chosen to focus primarily on non-ABM delivery channels and use the consequent savings to attract clients with lower service fees and/or higher interest rates on savings products. Indeed, some banks have made a strategic decision to provide services without branches.

These choices in the marketplace exemplify competition and provide consumers with considerable flexibility to have their individual needs met by their preferred financial services provider or providers. It would be important to consider, therefore, whether requiring institutions to expand functionality would give rise to cross-subsidization that would blunt this strategic differentiation and the competition it gives rise to.

b) Market Differentiation

Some banks take the view that ABMs represent one important way to differentiate themselves from each other and from their non-bank competitors in the financial services marketplace. They identify the following market differentiators:

Size of the Delivery Channels

– Each bank has taken its own strategic decision regarding the mix of branch/ABM/EFT/Internet and other delivery channels appropriate to meet the needs and demands of its clients and to attract new clients. As noted, some banks have invested more than others in building and expanding an ABM delivery channel because they have determined that having more ABMs available to clients is what will make their bank more attractive and competitive in the market.

Functionality of ABMs
– Many banks have made distinctive choices and introduced different innovations into their respective ABM delivery channels. While all ABMs dispense cash and most accept proprietary deposits, other functions differ from bank to bank depending on the bank's competitive strategy and the capabilities of its system architecture. Functions available at some bank ABM delivery channels and not others include cheque ordering, accessing $USD accounts and dispensing $US dollars, updating pass books, making mortgage inquiries, ordering tickets, and various methods of making bill payments.
Location of ABMs
– Where ABMs are located raises two observations. First, the majority of ABMs, approximately 70 percent, are located on branch premises. Historically, most clients demonstrate a strong preference for the flexibility and convenience offered by access both to in-branch and ABM services. Many banks have responded by making ABMs an extension of their respective branch networks, enabling clients to conduct as much of their banking at an ABM or in a branch as suits their individual needs and preferences. Second, banks have made competitive decisions regarding the location of their off-premises ABMs. This is generally done through negotiations with third parties, each of which has made a strategic decision to partner with a particular bank in permitting that bank's ABMs on its premises. For its part, a bank chooses its third party partners on the basis of client convenience and what it believes will provide competitive advantage.

Other deposit-taking institutions, notably those with fewer ABMs and smaller investments in ABM technology, state that market differentiation represented by ABM networks is no longer as important as it may once have been. They take the view that ABMs represent, in effect, a commodity, and that in their experience there is no evidence that consumers' choice of a financial service provider is based on the number, location or functionality of ABMs. The experience and views of these deposit taking institutions with smaller investments in ABMs is significantly different from those with substantial investments.

It is clear that there are differing views amongst banks – based to a large extent on the differing strategic visions of individual institutions – regarding the role ABM networks play in relation to an institution's business planning and to the competitive landscape. Policy makers will need to carefully assess whether or not market differentiation through proprietary ABM networks is an ongoing factor affecting competition in the marketplace. More specifically, policymakers will need to consider the implications of requiring a broadening of ABM functionality for those deposit institutions that view their ABM network as a proprietary source of competitive advantage.

3. Customer Service

ABMs represent an important financial service delivery channel for Canadians. In fact, they have become the primary means of conducting financial transactions for over 40 percent of bank customers. Accordingly, the CBA believes it will be important to assess the potential impact of broadening the shared functionality of ABMs on customer interests.

On one level, expanding ABM functionality would provide customers with more ABMs in the short run through which to conduct their banking services, and some observers point to this aspect of consumer convenience as a benefit that would arise from policy changes in this area. Others, however, believe this short run phenomenon will be reversed in the long run as forced commoditization results in a disincentive to investment and innovation. As well, there are other factors and potential implications for customer service that policy makers should also consider, and these are set out below.

a) Innovation

Canada's financial services marketplace is notable for introducing new technologies that provide greater convenience, choice and service for customers. These technologies stimulate innovation, enhance competition, improve business efficiency, and create new delivery channels and product offerings for the benefit of Canadians.

ABMs illustrate improved service and convenience through innovation. The number of services available to customers through ABMs has increased significantly since ABMs were first introduced over two decades ago. Innovations include bill payment services, obtaining information about mutual fund investments, ticket ordering functions, making mortgage inquiries, and special technologies improving accessibility to ABMs for disabled Canadians. The installation of ABMs at branches and the subsequent development of the Interac network for shared cash dispensing was a cornerstone to further development of new alternative delivery channels for banking services. With consumers quickly accepting access to cash at thousands of locations across the country, banks were able to build technology such as telephone banking and Internet banking that allowed their customers the ability to do their banking using other technologies at a time and place that was convenient to their customers.

In light of the importance of innovation, some deposit institutions have expressed concerns about the impact that mandating expanded functionality may have on innovation and have raised a number of questions. Policy makers may wish, therefore, to consider the potential impact of mandating shared ABM functionality on the incentive of financial services providers to invest not only in existing system but also future innovations. For example:

Would providers of ABMs add innovative new functions or accessibility features to their respective networks if the benefits of those investments were to be shared with non-customers? Would there be an incentive for a provider not to innovate, if that provider could simply choose to wait for other players in the market to innovate and then take advantage of that other provider's development?
If mandating expanded functionality did result in disincentives to innovation, could pricing of shared services be used to offset or mitigate those disincentives? In other words, would sufficient flexibility be allowed for providers to set prices in accordance with their individual assessment of costs and market conditions? It is useful to note in this regard that surcharging and convenience fees are not permitted in the Exchange Network. But, as noted below in our discussion of costs, increased costs would be proportionally greater if larger networks were to be required to expand functionality. The question arises what impact increased costs would have on innovation particularly where there would be no profit incentive to innovate?

Other observers, however, suggest a different approach to encouraging innovation. The Exchange Network, for example, distinguishes between common shared services (i.e., withdrawals, deposits, transfers between accounts, and account balance inquiries) and what it views as "competitive services" (e.g., bill payments), with the implication that individual institutions would continue to have competitive incentives to pursue their individual strategies and innovations in these latter areas.

b) Future Deployment of ABMs

Related to the question of future innovation is consideration of the effect of broadened shared functionality on incentives to expand deployment of ABMs. The primary issue to consider is whether the net effect of broader shared functionality would be to expand or contract the number of ABMs available to customers (or if it would have a neutral effect). Questions to consider include:

Would older machines be upgraded to incorporate added functionality or would they be eliminated and not replaced because customers could rely on the availability of others' ABM networks?
Would ABM providers have an incentive to remove, or conversely to expand, their own ABMs from off-premises locations in cases where customers of competitors became the major users?
Given the close linkage noted above between ABMs and bank branches, would off-site ABMs be particularly affected by decisions to expand or contract ABM networks? If so, what impact would this have on third parties involved in contractual arrangements regarding the deployment of ABMs on their premises?

c) Costs

It will also be important to consider whether expanding shared ABM functionality would have an impact on costs to consumers. As noted, members of the Exchange Network have decided to make its various shared services available to consumers without surcharging/convenience fees. Nonetheless, any cost increases associated with mandating shared functionality would have to be paid for, whether directly or indirectly. Indeed, there are a range of costs associated with ABMs, including the cost of installing, operating, and maintaining the machines; costs of upgrading software; and costs associated with location (e.g., are new machines being added to existing ABM locations, or is a new facility, requiring enclosures etc., being developed?) that need to be taken into consideration. These costs vary among institutions, and indeed, the ability to contain and manage the costs of providing ABM services is considered a source of competitive advantage by individual institutions. Moreover, plans to alter the size of an ABM network or to expand the functionality of a proprietary network would differ from bank to bank and would depend in large measure on the strategic business objectives of each bank. Thus, if shared functionality were to be mandated, each bank would need to assess its own proprietary costs associated with fitting within such a network.

If banks were required to add expanded shared services, they would almost certainly face additional costs associated with new software required to deliver services on a shared basis. Moreover, shared deposit taking would lead to large and/or unpredictable volumes of cheque and cash deposits being received in ABMs from non-customers. Accordingly, banks with extensive ABM networks would have to increase staffing levels or contract with third party providers to handle the increased volume of paper. Although many Exchange members report that they have not experienced the need to add staff to address this problem, volume levels in the Exchange are fairly low (e.g., one member indicated that only 60 shared deposits per month are received). Volumes are considerably larger, however, in the networks of larger banks and the impacts on staffing requirements correspondingly greater (see below for a further discussion of the implications of shared deposit-taking in this regard).

Additional revenues generated from surcharge fees might offset a portion of the added costs. But it is not certain that additional revenues would offset all of the costs associated with broadening shared functionality. Surcharges, on a full recovery basis, applied at the transaction level would likely be unattractive to customers. Thus each bank would have to assess the extent to which adding shared functionality might require cross-subsidization from its customers of other services and delivery channels.

4. Risk Management

a) Fraud and Money Laundering

It will be important to consider whether expanding shared ABM functionality may have implications for the financial sector's fight against fraud and money laundering. Fraud is an increasingly serious challenge to the financial services industry, and indeed the number of banking fraud cases has increased considerably in recent years. It is open to question whether permitting persons to use all ABM machines for deposits would lead to an increase in the number of empty envelope 'deposits' and other fraudulent activity across the system. It is interesting to note in this regard that, during the CBA's research into the operations of the Exchange Network, while some members interviewed did not experience an increase in fraudulent activity or perceive a concern one member of that network did note that fraud could definitely increase as a result of the expansion of shared services. Further analysis is needed in this area and, if the analysis demonstrates an increased risk of fraud, it would also be important to examine the steps that may be taken to mitigate fraud and their implications for customer service. For example, impacts on hold policies, an area of considerable customer concern, would have to be considered, as would operational efficiencies (e.g., disputes between banks with respect to loss allocation), and increased loss experience (a bank's own "know-your-customer" diligence would not protect it against customers of other Issuers that might have different customer acceptance standards).

Furthermore, the possible impact of broadening shared ABM functionality on money laundering activity requires careful consideration. Increasing the number of access points might facilitate the ease with each money laundering could occur. Therefore, the implications for, and the roles and responsibilities of, banks for reporting suspicious transactions would have to be thought through. In the Exchange Network, responsibility for reporting any transactions under the Proceeds of Crime (Money Laundering) and Terrorism Funding Act lies with the bank that issued the access card to the customer involved in the suspicious transaction.

b) Operational Risk Considerations

A related issue to consider is whether a requirement to offer shared deposit taking among competitors' ABM networks would alter the operational risk of participants so as to require additional capital charges to address incremental risk of unexpected losses.

A first step is to assess what operational challenges may impact Issuers (institutions that issue cards to consumers) and Acquirers (institutions that accept deposits on behalf of Issuers' cardholders) if functionality is expanded. In most cases, card Issuers would bear the preponderance of liability for ABM network transactions. However, Issuers would also be in a position to control the level of risk they wish to assume for "on-us" deposits (through their own ABMs) and "not-on-us" deposits (by their cardholders at another institution's ABMs) for shared deposit transactions. If Issuers assume greater risk, to the extent that such risks are predictable and quantifiable, the risk/reward parameters could be addressed through pricing and acceptance limits. However, this calculation would be problematic if pricing or access parameters were regulated, thereby constraining the pricing for risk options available to Issuers. Issuers might also place time and dollar constraints on new customers and perhaps delay issuance of debit cards until a customer has developed a verifiable and trustworthy credit history.

To a lesser degree, Acquirers may face incremental risks to the extent that they cannot limit or obtain Issuer indemnification for loss or default (although it should be noted that the Exchange Network has protocols similar to the Canadian Payments Association to address this issue). Operational risks of expanded functionality may include greater possibility of fraud, unpredictable volumes, and therefore staffing requirements, as well as uncontrolled exposure to high-risk customers of other Issuers

If we assume that, in aggregate, the potential for fraud, breakdowns or disputes under a shared functionality regime may increase the operational risk faced by an institution, would more capital be required? To the extent that anticipated or expected losses are built into an institution's profitability model, it may be that no additional capital charges would be necessary. It would only be the extent to which there are, or might be, unexpected losses that additional capital might need to be set aside. The question of the impact of this on capital would be dependent on the business and risk model, and, as such, cannot be quantified in the abstract.

More broadly, another consideration may be the potential for loss of brand or franchise value that might occur if reputation or competitive advantages of an institution's ABM network were undermined in the manner discussed above. The definition of "significant operations" and "significant risk" in CDIC's Standards refers to anything that can create an exposure to or influence an institution's earnings, liquidity, funding, reputation or brand value – and in turn bear on its sound business and financial practices for capital management.

By contrast, the Basel II rules now in development would seem less likely to come into play under this scenario because they deal with direct losses that have or will impact the income statement. Under Basel, capital requirements are revenue generated, hence the more significant losses, the greater the capital requirements. However, unless significant losses are expected to be incurred as a result of shared deposit arrangements, the potential for significant capital implications (if any) under the Basel II regime would be difficult to quantify at this time.

On balance, even if a shared functionality regime gave rise to greater operational risk and the possibility of material unexpected losses that would have capital implications, presumably these could be addressed through an Issuer's fee structure. However, to the extent that pricing might be regulated or limited, this risk mitigant would not be available.

5. Institutional Operations

Broadening shared ABM functionality would necessarily involve operational changes at participating banks. It will be important to assess the cost and efficiency implications of these operational issues. Set out below is a discussion of these matters in the context of account transfers, balance inquiries, and shared deposits. In addition, this section reviews the different considerations that come into play when considering commercial (as opposed to individual retail) accounts. Finally, this section also includes a discussion of the operational issues related to bill payment, which appear to be qualitatively different than other possible shared ABM services.

The CBA drew on its research into the operations of the Exchange Network for the discussion below. It is important to note, however, that while the Exchange is useful as a guide to the types of operational considerations that would be involved in expanding shared functionality, the volume of shared transactions in the Exchange is very small, in comparison to the volume of shared transactions across the entire banking and deposit-taking industry. In addition, while the Exchange does have some large members, most of its members are regionally focussed and considerably smaller and less complex than the five largest banks in the country. In short, expanding the functionality of ABM networks across the banking industry would be, by orders of magnitude, a much more complex and complicated undertaking than the operations of the Exchange.

a) Operational Issues

A key consideration is the extent to which expanding functionality will impact operational efficiency and impose increased manual processing requirements. The key area of the review in this regard is shared deposits. While there would be system changes in order for an institution to participate in a shared network (e.g., changing transaction formats and proprietary systems for deposit acceptance and processing, modifying settlement and reconciliation applications, etc.), the more significant implication is the extent to which shared deposits would require manual processing. In this area, the Exchange Network's experience is useful to consider:

Every business day, the depository in an Exchange ABM is emptied of all deposit envelopes and the Acquirer (or an Acquirer's third party service provider) reconciles the contents of the deposit envelopes to the amounts of the deposit transactions. If there are any discrepancies in the amounts in the envelopes compared to the deposit transactions, Exchange rules require that the Acquirer contact the Issuer to advise of the discrepancy and the transaction is reversed. The Acquirer endorses any cheques in the deposit envelopes and clears these items with all other cheques through normal clearing processes. The endorsement stamp on the back of the cheque is necessary as it indicates that should the cheque be dishonoured, the cheque is to be returned to the Issuer institution. Also included in the endorsement is the name and routing information of the Issuer institution, along with the card number of the card that was used to make the deposit. If shared deposit taking were mandated, each Acquirer, or its agent, would require an endorsement stamp for each institution whose customers might deposit "not-on-us" cheques with the Acquirer. Where Acquirers received large volumes of "not-on-us" deposits in an expanded functionality regime the risk of clerical error, leading to disputes and losses, would increase substantially.
When a deposited cheque is for an amount of $5,000 or greater, but less than $10,000, the Acquirer must photocopy the front and back of the cheque, and fax it to the Issuer. Where the total amount of the deposit is greater than $10,000 the Acquirer must photocopy the front and back of all cheques contained in the deposit and fax this information to the Issuer.
The Issuer is also responsible for any reporting of transactions under the Proceeds of Crime (Money Laundering) and Terrorism Funding Act.
If a cheque that was endorsed by an Acquirer on behalf of an Issuer's customer is returned (e.g., NSF, stale/post-dated, body/figures differ, etc.), the Acquirer must also photocopy the front and back of the item, fax it to the Issuer and advise them that the item has been returned. This must be done on the day, or no later than the next business day after the returned item is received by the Acquirer through the clearing system, to ensure recourse is not lost.

At current volume levels, the Exchange processes for deposit reconciliation appear to require little, if any, additional staffing resources to verify and reconcile deposits. However, this would not likely be the case if all banks were required to provide shared deposit functionality using processes similar to that the Exchange Network. As noted above, Exchange volumes are relatively small, with one Exchange member indicating that the average number of shared deposits per ABM is 60 deposits per month. Extrapolating this across the entire CBA member ABM network could equate to approximately 998,000 shared deposits per month (although the actual number would depend on specific customer behaviour and willingness to use another institution's ABMs). If some or all of the above processes had to be put in place across the industry, the staffing requirements and impacts on operational efficiency could be significant. In effect, shared deposits would essentially graft a paper-based system to what is largely an electronic system. Further research and consideration would be needed to fully assess the impacts and changes that would be needed in this area.

By contrast, account transfers and balance inquiries in a shared network are primarily electronic, and would require only system changes (e.g., transaction formats, settlement and reconciliation applications, etc.) by institutions to be able to participate in a shared network and to "recognize", accept, and process transactions from other institutions in the network. Although there would be costs and operational impacts associated with these changes, they would not likely be as significant as those noted above. Depending on the network that would provide the shared services, there may also be costs related to system changes to the network to allow for additional shared functionality – further study would be required to determine what these costs would be, and, as noted above, this would depend in large measure on what the definition of expanded functionality would be.

b) Commercial Account Implications

Consideration of expanding ABM functionality gives rise to the question of whether the expanded functionality would apply to individual retail accounts only or to commercial accounts (i.e., small/medium-sized businesses). As noted below, commercial accounts would give rise to a range of additional issues that would need to be addressed.

At the outset, it is important to note that many banks in the normal course of business do not provide ABM service to their commercial accounts. In point of fact, members of the Exchange Network do not typically provide access to the shared services of the Network to their small business customers. This is due to the operational and risk issues set out in the sub-bullet points below. Moreover, it is open to question whether small business owners would choose to leave their deposits in the ABM of a financial institution that they do not deal with in the normal course of their business.

Mandating that commercial accounts share ABM functionality would raise numerous issues, including:

The potential for greater operating risk faced by Acquirers due to increased chance of errors and missing cut-off times due to heavier commercial volumes:
Current Exchange Network processes can accommodate manual handling on small retail deposits, for example the deposit of one cheque to a retail account. Handling multiple cheques deposited to a commercial account would require extensive and costly manual handling, including reconciliation of deposits to keyed-in amounts and the notification by the Acquirer to the Issuer of discrepancies. By way of example, and as noted above, the Exchange Network rules require photocopying and faxing of cheques over $5,000 from the Acquirer to the Issuer. The potential for receiving cheques over $5,000 would be increased considerably if commercial deposits were permitted. In this regard, one bank reports the average deposit for a commercial deposit is $7,100 as compared to $502 for consumer accounts. Similar cost issues arise in the context of reporting under the Proceeds of Crime (Money Laundering) and Terrorism Funding Act.
The potential risks of turning an ABM into a night depository for businesses, and requests by non-customers of the Acquiring institution for additional business services (e.g. small denomination bills, coin, remittances, etc.):
Although banks now accept cash deposits in their proprietary ABMs, the trend is away from such use by consumers for security reasons. The problems would be considerably magnified in light of the often-large volumes of cash that small businesses would need to deposit on a daily basis – in effect, ABMs are not a substitute for night depositories, which are specifically designed and built to accommodate commercial deposits.
In a related consideration, the physical size of commercial deposits relative to the deposit slot in an ABM:
The standard size of deposit slots in ABMs would restrict deposits to a thickness of no more than 20-30 items per deposit. Although some very small businesses may have deposits of this size, many commercial accounts would have regular deposits of considerably larger volumes.

c) Bill Payments

As noted, bill payments are qualitatively different than many other types of transactions that may be considered under expanded functionality and warrant a separate discussion. It is instructive to note in this regard that the Exchange Network does not include bill payments as part of its shared functions because members consider it to be a "competitive" function outside the scope of the four core shared functions. The government's June response paper, however, mentions the possibility of including bill payment as a component of broadening shared functionality. Therefore, some of the operational issues that might arise are set out below.

The processing of bill payments has evolved from the method where customers selected the bill payment option at the ABM, and inserted an envelope into the ABM with the bill payment stub for remittance. The system is now a fully electronic process completed on the ABM screen, eliminating the need to submit paper stubs. In effect, the ABM is one alternate delivery channel that customers can use to pay their bills, with others being telephone banking and Internet banking (along with paying a bill at a bank branch).

The number, mix, and geographic location of billers will be unique to each customer and this creates a greater challenge to incorporate bill payments into shared functionality (and a correspondingly greater challenge in adapting systems to incorporate this function). In a shared environment, when a customer selects the bill payment function, the request would have to be sent to the Issuer and the customer's bill payment portfolio would be sent to the Acquirer's ABM for the customer to select the appropriate bill(s) for payment. Current ABM hardware technology usually limits the number of bills that can be displayed to the customer to a maximum of ten companies at one time. If the bill to be paid were not in the first ten bills that the customer had registered, then another request would be sent to the Issuer for the next ten bills (and so on) until the customer found the appropriate bill(s) to be paid. This type of transaction adds time and network "overhead" to each bill payment transaction, and is considerably more complex than, for example, a shared balance inquiry where the customer identifies the specific account he or she wishes to access.

6. The International Experience

Other leading countries also have experience with shared ABM functionality. Although not an exhaustive survey, set out below is a summary of the experiences of other key jurisdictions often used as comparisons for the financial sector and the public policy framework in Canada (i.e., Australia, the UK, and the U.S.)

As is noted in more detail below, in neither Australia nor the U.K. is expanded or shared functionality mandated. While there has been a measure of mandated shared functionality in some states in the U.S., increasingly it is being superseded by efforts to foster competition in payment service systems as a preferred outcome.

a) Australia

Australian banks generally offer shared ABM functionality along the lines provided by the Exchange Network in Canada. Similar to the Exchange Network, this shared functionality was a market driven response rather than a legislated requirement. Individual banks have made an independent commercial decision to participate in a shared ABM network and have entered into bilateral and multilateral agreements to bring this into effect. The process followed in Australia for achieving expanded functionality is described by Australia's ATM Industry Steering Group (an industry stakeholder group) as follows: "Non-banks, because their own ATM networks were small, led the way in expanding the numbers of ATMs their cardholders could access by negotiating access to other institutions' ATMs." (Discussion Paper: Direct Charging for 'Foreign' Automatic Teller Machine (ATM) Transactions in Australia, October 2000)

b) The United Kingdom

The scope of shared functions appears to be more limited in the UK than in Australia. Similar to the Australian experience, these shared functions were not mandated by government. Until recently, the only shared functions of the UK system were legacy shared cash dispensing and balance enquiry functionality resulting from the absorption by LINK (the ABM system in the UK owned by banks and building societies) of other networks that provided these shared functionalities. More recently, in connection with a banking industry initiative to replace UK credit and debit cards with chip cards, the banking industry has put forward a new shared functionality – a pilot project allowing customers to change their PIN numbers at any UK ABM.

The UK payments system has been subject to considerable scrutiny in recent years, most notably through the Cruickshank Report of 2000. Despite the scrutiny, however, neither the Cruickshank Report nor the government's response to that report has proposed mandating shared functionality on the ABM network.

c) The United States

Mandated shared ABM functionality has been a feature in some U.S. states. A key factor behind mandating was the highly fragmented nature of the U.S. banking market, in contrast to the experience in Canada with large banks offering services nation-wide. Increasingly, however, state governments have been moving away from a mandating approach. Set out below is a summary of rationales for the early requirements for shared functionality, as well as reasons for the recent changes.

When ABMs were introduced in the U.S., they were proprietary machines used exclusively by a bank's own customers. By the mid-1970s banks had started to share ABMs on a voluntary basis, allowing other banks' customers access to their machines. At that time, a number of state governments instituted mandatory sharing laws that required any ABM-owning institution to share their machines for a "reasonable fee" with any other financial institution in the state that wanted to share. The rationales for such laws were ensuring access for customers and concern for smaller banks' ability to compete with larger institutions.

The experience of Illinois with shared ABM functionality is illustrative. Illinois mandated shared deposit taking from 1979 until the mid-1990s, when it made shared deposit taking optional. The context of Illinois banking in the 1970s suggests the original reason for mandatory shared deposit taking: when ABMs made their debut in the 1970s, Illinois law only permitted a bank to have one main office and two satellite locations at which the bank could take deposits and dispense cash. Therefore, to gain network effects Illinois law required banks to share ABMs, including the taking of shared deposits.

In the 1990s, however, Illinois dropped its restrictions on branch banking. By that time, a number of ABM networks had been introduced and the number of ABMs had increased substantially. With the proliferation of ABMs and competing networks, Illinois' rationale for mandating shared deposit taking had disappeared and the state made shared deposit taking optional. Once deposit sharing became optional, practical concerns and market considerations (discussed in more detail below) led to a reduction in the number of institutions willing to participate in shared deposit taking.

The CBA has been able to identify only a handful of states that mandate shared deposit taking today: Iowa, Michigan, Indiana and Wisconsin. While our review of state requirements has not been exhaustive, and while there may be states that we have not identified that mandate shared deposit taking, it appears that mandated shared deposit taking in the U.S. is on the decline.

As state governments enacted their own ABM network laws, the U.S. Congress sought a national response to the impact of ABM technology on banking and consumers. Congress created the National Commission on Electronic Fund Transfers in 1974, directing the Commission to advise what form of regulation, if any, was necessary to ensure a safe, fair and predictable EFT platform in the United States. One of the most contentious issues considered by the EFT Commission was state mandatory sharing statutes. While the Federal Reserve proposed mandatory sharing, the Justice Department opposed it, concerned that banks would be deterred from forming networks because they would have allow others to use the networks that they had invested in developing.

The debate centred around two views of competition policy: Are ABM networks a natural monopoly that should be regulated like public utilities? Or, should numerous networks be allowed to compete with little or with light regulation, and let market forces determine the range of services? The EFT Commission concluded that ABM networks were not natural monopolies and that the competing network model was the preferred policy option. Although Congress considered legislation to pre-empt state sharing statutes, it did not enact it. Rather, what unfolded was vigorous inter-network competition. Furthermore, the Justice Department's concerns about the anti-competitive effects of mandatory sharing appear to have been realized. In states that mandated ABM sharing, deployment of ATMs and card usage were lower than in states that did not mandate ABM sharing.

Today, mandatory shared deposit taking in the United States appears to be coming to an end, both for policy and operational reasons:

Policy Considerations
– Iowa, a state that continues to mandate shared deposit taking, extends the sharing requirement to off-premises machines only. On-premises machines are viewed as part of the branch delivery channel and not subject to mandatory sharing. Moreover, in asserting the paramountcy of federal jurisdiction over state banking law with respect to nationally regulated banks, the courts have entered the shared functionality debate. As part of its shared functionality law, Iowa prohibited banks from surcharging non-customers who used their machines. In 2002, five nationally chartered banks, with the support of the United State Office of the Comptroller of the Currency, succeeded in overturning Iowa's 24 year-old surcharge ban. The court affirmed the paramountcy of U.S. federal banking law that does not require shared functionality.
Operational Considerations
–With the end of mandatory shared deposit taking in Illinois, Bank One stopped taking shared deposits in its Chicago area ABMs. The reasons cited by Bank One were:

Increased costs of shared deposits.
Transaction adjustment and settlement issues.
Increased fraud.

More recently ABM networks have changed their contractual sharing rules pertaining to shared deposit taking. On September 1, 2002 the NYCE network permitted its member banks to surcharge non-customers for deposits up to US$2.25 per deposit. NYCE network fees are similar in amount to other networks that offer shared deposit taking. Commentators note that higher interchange fees for shared deposits better reflect the cost of supporting deposit taking ABMs. On July 1, 2003 the NYCE network went further, making shared deposit taking optional for its participating banks. Networks and their bank members cite declining demand for shared deposits as their reason for no longer mandating shared deposit taking. Falling demand is attributed to:

Lengthy clearing times of up to five days for 'not-on-us' deposits.
The high cost to financial institutions of handling 'not-on-us' deposits and resultant customer avoidance of surcharge fees levied by financial institutions to recoup high operating costs.
General reluctance of customers to place deposits in foreign machines.
A decline in demand for shared deposit taking attributable to the increase of direct deposits by employers, health and dental insurers, and government.

7. Approaches to Achieving "Shared Functionality"

If, after its consideration of all the issues, the government were to decide to mandate expanded ABM functionality, this could be done on either a sector-wide basis or a selective basis. Both possible approaches, and their respective policy implications, are examined below.

a) Sector-Wide Approaches

The federal government could mandate shared functionality in two ways. First, Parliament could pass legislation directing federally regulated deposit taking institutions offering ABM services to expand shared ABM functionality. Second, the Minister of Finance could use the powers under The Canadian Payments Act to designate a payments system and issue a directive that the designated system must adopt shared functionality.

Mandating broader shared functionality across the financial sector would be a significant break from previous practice and would introduce a public utility mind-set that could have important ramifications for the future direction of financial services policy in Canada. Among other things:

It would shift the status of the ABM delivery channel from that of a competitive, market-driven one towards being an instrument of government policy akin to a public utility. Consequently, this could result in considerable uncertainty in the marketplace and act as an inhibitor of future innovation. Questions would arise about which future directions the government may give to the system and how consistent these directions might be with individual financial institutions' strategic objectives. Having specific regard to designation, there could be a disincentive to make further investments, whether to replace older machines, make upgrades or add innovations, since costs would be unpredictable as a result of increased uncertainty about which system rules the Minister would or would not approve or amend in future.
Such a step could have implications for how Canada is viewed on the international stage. If international institutions were to form the view that there is a growing degree of government involvement in the direction of the financial system, it could act as a disincentive to future investment in this country. Accordingly, such a decision would need to be considered from the perspective of whether it would retain and promote investment in Canada or encourage the migration of capital and creation of financial centres in other jurisdictions.
Designating a system to offer shared functionality could have significant competitive implications for non-designated ABM networks.
Mandating would raise the issue of whether third-party white-label ABMs would also be required to offer expanded shared functionality, as well as the limited functionality ABMs that some institutions have as part of their ABM portfolio.
The operational and other implications arising from adding shared functionality identified throughout this submission would need to be carefully considered.

b) Selective Approach

One selective approach would be to make shared functionality a condition for approving a merger between large banks (i.e., merger proponents would be required to broaden the shared functionality of the ABMs of the new merged institution). This approach raises a number of considerations.

One of the first issues to be considered would be whether or not the merged bank would undertake to seek shared ABM functionality through bilateral or multilateral arrangements within Interac or be required to join the Exchange Network. With respect to Interac, the following four conditions would apply to an attempt to any consideration of new bilateral or multilateral shared services:

The service must not replicate an existing Interac service provided to members.
The service must not result in any material adverse impact to Interac.
Market viability of the service must require real-time access to member demand deposit accounts.
The Interac members participating in the bilateral or multilateral service must bear the costs of developing the new service and any consequent changes required to the Interac Inter-Member Network.

Having regard to requiring banks to become Exchange Network members as a condition of approving a merger, the Exchange Network is currently set up so that transactions between two Exchange members bypass the Interac system. Thus, should an Interac member bank be required to join the Exchange Network, all future 'not-on-us' transactions involving other Exchange members, including withdrawals, would go through the Exchange, not Interac. This would result in lower revenue for the Interac system. Accordingly, policy makers would need to assess the operational and revenue implications for the Interac system. Similarly, if Interac were to be designated, policy makers would need to assess these implications for the Exchange Network.

Moreover, if broadening shared ABM functionality were to be a condition of a large bank merger, a question that would need to be considered is what the scope of that shared functionality would be.

Would all deposit-taking institutions have access to the merged banks' ABMs and, if so, would the merged entity have the right to expect access to all other deposit-taking ABMs? If so, the merger-condition approach would mean that functionality would be extended across the board.
Or, would only smaller financial institutions have access to the ABMs of the merged bank, and if so, on what basis, e.g., would it be limited to the members currently participating in the Exchange Network? In the latter case, customers of non-merging members of Interac who are not also members of the Exchange would not participate in the broadened functionality system. Such an approach would allow some banks (as well as credit unions) to make use of the large ABM infrastructure of the merged bank, but would not allow that access to non-Exchange members. Since the former category includes both small and relatively large banks (as well as regional banks and banks with national scope), it is not clear what public policy goals would be achieved through the resulting asymmetry in the marketplace.
Or, would the merged bank be required to share its ABMs with any financial institution that expressed an interest in having access, and if so, would the sharing be on a mutually agreed contractual basis or on terms set as a condition of the merger? This would raise the question: through which shared network would the shared functionality be achieved and how, in practical terms, would that be achieved? Would, for instance, the institutions interested in sharing with the merged bank be the same as the membership of the Exchange and if not, would that raise operational or governance issues within the Exchange? If the intent were that it would be achieved through the Interac system, how would that be accomplished of only some Interac members were involved?

8. Other Policy Considerations

a) Investment

Over the years, many banks in Canada have made very large investments in their proprietary ABM network. Although banks voluntarily agreed to share certain services through Interac, many of them continue to view ABMs as a competitive market differentiator and allocate capital to this delivery system accordingly. Any decision to require institutions to add to the functionality of their machines in a way not in keeping with their individual business plans, and to share these machines with other banks, trust companies and credit unions, would need to take fair and careful consideration of compensation for the investments made by individual institutions.

b) Precedential Implications

Given the evolution of the retail marketplace, the ABM network is now seen as a one of a range of delivery alternatives to consumers, comparable in many ways to the branch network (and other delivery channels). If the entire ABM network were required to be expand its functionality, it could be reasonably argued that this would set a precedent for requiring that shared services be offered in the branch network (particularly where ABMs are located in branches) and may in fact be viewed by some stakeholders as a logical next step. Such a step, of course, would represent a fundamental transformation of the branch network and impose a dramatically different business model and rationale for building and maintaining branches than is now in place.

It is sometimes observed that the shared withdrawal function at ABMs did not lead to shared withdrawals in branches, and that therefore there would not be a spill-over effect if functionality were extended. It should be noted, however, that sharing withdrawals through Interac was a voluntary decision made by Interac's members. If the government were to require expanded functionality of all ABMs, a matter to be considered is whether the functionality of the system would no longer be a voluntary matter, but rather a matter for government policy direction. In this scenario, it would be a much smaller step for the government to move from mandated shared functionality in one delivery system to mandated shared functionality in another, than to move from voluntary shared functionality on a single type of transaction to shared functionality within branches (or other delivery systems).

c) Voluntary vs. mandatory

At the heart of the discussion is whether ABM functionality should continue to be a voluntary matter for individual institutions to consider, based on their individual business plans, or whether expanded functionality is to be required of institutions by the government.

Regarding the latter choice, it is sometimes observed that government action to mandate shared functionality would be analogous to the 1996 Consent Order regarding the Interac network. There are important differences between the two situations, however. The 1996 Consent Order required the owners of the Interac system to allow others to join the system on terms established through the Consent Order – it did not change the Interac system itself or require new functions to be added to the system. Mandating ABM functionality, on the other hand, would require financial institutions to add new services to their ABMs and incur new costs in doing so.

Voluntary shared functionality is the approach taken in Australia, the UK, and increasingly in the US, and is the current approach in Canada through the Exchange. It is based on the premise that individual institutions can follow their own strategic business plan in choosing whether or not to offer shared functionality. To change this model, i.e., to move to a system where the services available through a delivery channel are determined by government policy, would be a significant break from previous practice and could have important ramifications for the future direction of financial services public policy in Canada.

9. Conclusion

The question of broadening shared ABM functionality in Canada provokes many significant public policy issues. A shift in current policy could have significant implications for the future direction of the Canadian banking industry, non-bank competitors, competing ABM networks, and for consumers of financial products and services. Policy makers should give careful consideration to each of the issues and questions we identify in this canvass of the possible impact of expanded shared ABM functionality on competition in the marketplace, customer service, risk management, and institutional operations.

We trust that the banking industry's experience in operating ABM networks as discussed in this submission will be of assistance to government policy makers.